🤖 Sam Outman

🧵 In today's edition:

• 🤖 Sam Outman

• 🔐 Meta disbands the Responsible AI team

• 🎥 Surveillance GPT

• 🎙️ New Episode ft. Paul Yacoubian, CEO at Copy.ai

• 🤑 AI Fundraising News

• 🗞️ AI Quick-Bytes

• 🐦 Tweet Post of the Day

🤖 Sam Outman

In the craziest plot twist of all time, Sam Altman was fired from OpenAI.

Here’s a quick timeline of Friday’s events (Stick with us - because this reads like a future Oscar-winning movie script):

• At noon on Friday, Sam Altman was informed by Ilya Sutskever (board member) that he was being fired, and Mira Murati became the Interim CEO.

• At 12:23 pm, Greg Brockman, President and Co-founder of OpenAI, learned about the change in leadership from Ilya Sutskever.

• At 12:25 pm, the public got the news through a statement.

• Sam confirmed his firing on Twitter.

• A few hours later, Greg Brockman quit, expressing disappointment in the board.

When the news broke, the Silicon Valley/tech community flooded Twitter with love and support for Sam, demanding the board's removal and Sam's return.

Just 24 hours later, The Verge reported the board wanted Sam back, but he'd only return if they resigned.

Overall: Sam might be able to pull off a Steve Jobsian-type comeback to OpenAI within 72 hours.

But here’s the thing - we need to figure out why this happened. OpenAI's goal is clear - they're a non-profit aiming to create safe and responsible AI. Yet, the board fired the company's creator with a vague statement that sounded like typical corporate talk, with no explanation of what actually took place or any evidence that Sam went against the company’s goal.

I'm not saying the OpenAI board is predicting doom, but this news will be intriguing to follow all week. We'll keep you updated :)

Read more: Engadget

🔐 Meta disbands the Responsible AI team

Meta announced that it’s disbanding its Responsible AI (RAI) team to focus more on developing its GenAI infrastructure.

For context, the RAI team was established in 2019 to address issues in AI training, preventing moderation problems on Meta's platforms. 

The good news: No one is being fired. Instead, most of the group is being moved to Meta’s GenAI product team.

The concern: How does this move affect Meta’s AI safety governance?

The company has assured users and critics that they will continue prioritizing responsible AI and that the restructured team will support cross-Meta efforts in this domain, but we are yet to see what that looks like, process-wise.

Why the change? This decision follows a prior restructuring that Business Insider described as leaving RAI "a shell of a team," highlighting autonomy and stakeholder negotiation challenges. 

Overall: Even though Meta doesn't have a separate team just for RAI anymore, people are hopeful that Meta will still focus on making sure its AI systems are safe and secure. 

This is necessary, especially at a time when countries are working on rules for how AI should be developed as worldwide adoption increases.

Read more: The Verge

🎥 Surveillance GPT

Forbes recently reported that ChatGPT is moonlighting as a surveillance tool for a Russian spyware company, Social Links. 

Social Links has been using ChatGPT for sentiment analysis, which involves analyzing social media sentiments to gauge how internet users feel about various topics.

But they aren’t exactly hiding their project. After being previously ousted from Meta’s platforms for alleged snooping, Social Links showcased its unconventional use of ChatGPT at a Paris security conference. 

How did they do it? The chatbot's text summarization and analysis abilities were weaponized to sift through extensive data quickly to conclude how people react to different things. 

As suspected, privacy advocates are beyond alarmed, not just about this specific case, but for the broader implications of AI empowering the surveillance industry.

Overall: The use of AI in such scenarios raises concerns due to chatbots' track record of errors. AI training data has historically been a "black box,” making the algorithm's responses unreliable, especially when trained on platforms like Reddit and 4chan, which are known for biases. 

This revelation shines a light on how AI might be misused in important situations, warning us to be careful when using it in more serious cases beyond just casual chatting.

Read more: Gizmodo

🤑 AI Fundraising News

• OfferFit raises $25M in Series B funding to create a machine learning-based solution to optimize and personalize marketing efforts, challenging the effectiveness of A/B testing in digital marketing.

• Cytovale raises $84M in Series C funding to create a diagnostic test called IntelliSep, which assists in the early detection of sepsis.

🗞️ AI Quick-Bytes

What else is going on?

• Actors are concerned about AI terms in the new SAG-AFTRA contract.

• Nvidia defies spending slump weighing on Best Buy and HP.

• OpenAI has received just a fraction of Microsoft’s $10 billion investment.

• Do Pixel 8 AI-edited photos have to be labeled as synthetic content on YouTube? YouTube is still figuring it out.

• WhatsApp adds a cute lil' button that lets you chat with Meta AI.

• Amazon is laying off several hundred employees working on Alexa.
  
  

🐦 Tweet Post of the Day

😂😂😂